Integrated steel manufacturing processes heavily rely on steelmaking gases. These act as fuel for crucial downstream processes like electricity generation, reheating furnaces, sintering, pelletizing, and calcination. Optimizing the distribution of these gases presents a significant challenge for integrated sites. The dynamic and complex nature of gas usage leads to energy waste through excessive flaring, increased costs, and heightened carbon emissions, exacerbated by a lack of visibility and transparency between isolated production units.
Integrated sites recover and use Blast Furnace Gas (BFG), Coke Oven Gas (COG), and Converter Gas (BOFG) in downstream processes. These gases are mixed or enriched with natural gas at mixing stations based on consumer needs like heating value and composition. Boilers produce steam for use across the site at various pressures. Gasholders offer limited storage due to high flow rates. When supply exceeds demand, excess gases are flared, and extra steam is vented. Optimizing gas and steam flow rates and mixing ratios at integrated sites is usually suboptimal and needs improvement.
Viridis Dispatch addresses these challenges with a non-intrusive approach, targeting the optimization of steelmaking gas and steam distribution in integrated plants. Leveraging sophisticated artificial intelligence methods such as machine learning and evolutionary algorithms, this powerful tool simulates and optimizes different operational scenarios to determine optimal dispatch process parameters and setpoints. It seeks to minimize flaring, maximize downstream equipment performance, and increase energy generation, all while considering operational and safety restrictions.
Customer challenges addressed:
> Dynamic and complex nature of gas usage leads to energy waste through excessive flaring
> Increased costs due to inefficient gas distribution
> Heightened carbon emissions caused by suboptimal gas balance
> Difficulty in improving gas and steam distribution parameters manually
> Operational and safety restrictions affecting gas distribution optimization efforts
Key features:
Viridis Dispatch addresses industry challenges in data collection:
Viridis Dispatch maximizes efficiency by leveraging the Viridis Suite and industrial ready-to-use connectors, enabling seamless data collection from various sources within industrial complexes. It also offers an Software Development for expanding connector options and includes built-in data cleansing functionality.
Continuous improvement in gas consumption forecasting with real-time production order adjustment:
One of the standout features of Viridis Dispatch is its robust capability to continuously enhance gas consumption forecasting, even during production order adjustments and rescheduling. This adaptability ensures that forecasts remain accurate and reliable, accommodating the dynamic nature of industrial production environments. By integrating real-time data and sophisticated algorithms, Viridis Dispatch can swiftly recalibrate forecasts to reflect changes in production schedules, thereby optimizing gas usage and reducing wastage.
Optimize Gas and Steam Distribution with Advanced Forecast Models:
The first step in running an optimization scenario is evaluating all configured forecast models, which typically include variables such as the generation and consumption of steelmaking gases (coke oven, blast furnace, melt shop/converter) and steam. These models are configured and trained using multiple algorithms, historical data, and contextual information, and can be individually defined and tuned or based on Viridis AutoML functionality. The output of the forecasting process is a set of time series used in the dynamic simulation, ensuring precise and efficient gas and steam distribution.
Enhance operational efficiency with advanced optimization and simulation:
Viridis Dispatch features configurable multi-objective optimization and dynamic simulation mechanisms. These mechanisms allow users to set decision variables such as gas recovery (burning in flare or not), gas turbine consumption, the number of burners activated (and the boilers' gas consumption), and the type and proportion of gas mixtures in mixing stations. Objectives include maximizing electricity generation in turbines, minimizing natural gas and other support gas consumption, reducing burner commutation for stable boiler operation, and minimizing emissions and gas burning in flare.
Based on configured parameters and previous runs, the optimizer generates a list of candidate solutions. For each candidate solution, the optimizer triggers the dynamic simulation, which produces time series data for each simulated metric (e.g. gas holder level over time), scores for each optimization objective function (criterion), and feasibility scores for each operational restriction. The optimizer then selects the best solution by weighting the scores for each criterion, defining the setpoints for optimal performance.
Efficient control with recommendation and closed-loop modes:
Viridis Dispatch defines setpoints for control variables over the planning horizon, such as the next 60 minutes. It operates in two modes to suit different operational needs:
- Recommendation mode: setpoints are provided to operators, who can choose whether to follow them. The system tracks adherence, offering an audit functionality to monitor operator compliance.
- Closed-Loop (Actuation) mode: setpoints are automatically sent to automation systems via the Viridis IoT gateway, ensuring real-time, hands-free optimization.
Tailored Dashboards and Comprehensive Scenario History
Viridis Dispatch offers the flexibility to configure multiple operational views based on customer needs. These views are supported by configurable dashboards, providing users with tailored operational perspectives. Additionally, Viridis Dispatch maintains a complete history of all scenario runs, allowing users to select and review past runs to analyze performance and outcomes.